TSR2

by Damien Burke

  Buddy refuelling

  Included in the initial requirements was the capability for any TSR2 to act as a buddy refuelling tanker for other aircraft. This would be achieved by fitting a buddy refuelling pack within the weapons bay, with the bay doors removed and a bulbous fairing housing a hose-and-drogue unit (HDU) based on the Mk 20 unit then in common use with Fleet Air Arm Scimitars (the TSR2 unit was to be the Mk 26). Unlike the Mk 20, which generated its own electrical power via a small wind-driven turbine, the Mk 26 was powered by the aircraft’s own AC system. An arm would lower from the fairing to give the hose and drogue sufficient clearance away from the hot exhaust gasses from the engines and also the vortices generated by the wings. Work began on this pack at Flight Refuelling Ltd in 1960, and progressed fairly well through 1961. The capacity of the pack itself was not overly impressive, being a mere 400gal (1,818L) (reduced to as little as 325gal (1,475L) when allowance was made for the channel into which the extending arm would retract). With the pack closed, the aircraft’s performance was required to be the same as with no pack fitted at all, though lower g limits would apply. With the boom arm deployed, a 450kt (520mph; 834km/h) limit would be in place (the same as for flying with the aircraft’s own refuelling probe extended).

  A general arrangement of Flight Refuelling Ltd’s Mk 26 buddy refuelling pack. BAE Systems via Warton Heritage Group

  In August 1962, after the RAF had carried out a general survey of planned TSR2 operations, it decided that it could do without buddy refuelling, and the requirement was promptly cancelled, saving £0.5 million in the process. The ability to receive fuel from Victor tankers or other types was, however, still a requirement. A study comparing the TSR2’s flight envelope with that of the Victor was carried out by BAC, and it appeared that throughout the band in which Victors carried out refuelling, the TSR2 would be able to plug in and keep up in dry power.

  Overload tanks

  Provision for additional fuel tanks to extend the aircraft’s range was written into the requirement and specification from an early date, but actual work on this was very much an on/off affair, with apparent bursts of enthusiasm for the idea punctuated by months of inactivity. Overload tanks to be carried under the wings were expected to be of 450gal (2,045L) capacity each. As the wing was to be designed for external stores carriage on a pair of pylon positions, the first suggestion was to carry drop tanks on one or the other of the proposed pylon locations. However, windtunnel tests with early stores layouts had shown that there were significant destabilizing effects from larger stores and a 450gal fuel tank certainly came under that description. After very brief consideration of an overwing location similar to the overwing tanks on the Lightning, a slipper fuel tank attached directly to the wing was drawn up.

  Flight Refuelling’s Mk 26 buddy refuelling pack reached mockup form before the requirement was dropped. BAE Systems via Brooklands Museum

  The 450gal (2,045L) slipper tank design originally favoured. A pylon-mounted drop tank became the preferred underwing tank later in the project. Damien Burke

  This tank was to be positioned under the wing on a line centred at station 120, the inner pylon mount point. This 450gal tank would afford extra endurance but no actual increase in combat radius unless it was jettisoned when empty (indeed, retaining the tank would actually decrease combat radius); definitely a wartime or emergency-only exercise owing to the expense involved. Jettison would be achieved by an ejector gun pushing the tank downward at its centre point, rotating about the rear attachment until it reached 90 degrees, at which point the rear attachment would disengage and allow the tank to fall free, clear of the tailplane. Windtunnel tests without an ejector gun had found that the tank remained in place or would not drop away cleanly in certain flight attitudes, particularly at slow speeds with the aircraft pitched up (such as during take-off and landing).

  The slipper tank later fell out of favour, and pylon-mounted drop tanks re-entered the picture. With a rounded nose similar to the slipper design, the new tanks would not be cleared for supersonic flight, but they were aerodynamically superior to the slipper tanks, and it was estimated that the decreased drag would mean that they would now carry no range penalty if retained when empty, and on the long-range ferry sortie profile a small range increase would result even if the tanks were kept. Jettisoning them when empty was still the only way to gain any appreciable benefit on a combat sortie, however, and in this case an additional 200nm (230 miles; 370km) range (100nm combat radius) would be gained. The fuel cock permitting transfer from the drop tank to the wing tanks was a modification embodied on all TSR2 airframes from XR220 onwards; XR219 would have been unable to carry drop tanks as a result.

  A variety of schemes for a fuselage overload tank were also looked at. Initially, in 1961, just two versions were investigated, a 570gal (2,590L) weapons-bay tank (held entirely within the bay and giving 250nm (290 miles; 465km) extra range) and a 900gal (4,090L) tank (which would protrude below the bay to about the same extent as the reconnaissance pack, and require the bay doors to be removed, but give 400nm (460 miles; 740km) more range). In July 1963 a feasibility study into the provision of a jettisonable ventral tank was completed. It outlined a pair of designs, one of 1,000gal (4,545L) capacity and one of a whopping 1,425gal (6,475L). These were rather more attractive than underwing tanks because they would still give a sizeable range increase even if retained when empty. The smaller of the two gave an additional 280nm (322 miles; 520km), and the larger 390nm (450 miles; 725km). Jettisoning the tanks when empty would increase these figures to 400 and 560nm (460 miles; 740km and 645miles; 1,035km) respectively. These tanks would also be supersonic, unlike the wing tanks. Minor structural changes would be necessary on the airframe to cater for the ventral tank attachment points, and the ventral anti-collision beacon would need to be moved, but the ventral ILS aerial would not need to be moved as the rear part of the tank covering this area could be made of glassfibre.

  In May 1964, after the arrester hook requirement had been raised, the lines of these tanks were altered so that they could be compatible with the expected hook installation, and the RAF also began to take a long hard look at just what it needed in the way of overload fuel provision. The wing tanks, despite their disadvantages, left the weapons bay free for its intended purpose and would be relatively cheap and fast to develop, so they were in. The internal weapons bay tank was also definitely needed, for ferry purposes. The larger ventral tanks would be more expensive to develop, and introduced possible problems when used in combination with internal weapons. In addition, the engine accessories bay was desperately in need of a redesign, and this could possibly have an impact upon the ventral tank lines, so the big tanks were put on the back burner, to be looked at in the future. The combination tank that sat half in the bay and half out was of no interest, and was dropped.

  By the end of 1964 the cautious attitude at the Air Ministry and MoA, where the project’s future was already felt to be in serious doubt, meant that no further work on overload tanks was authorized before the cancellation.

  A model of the pylon-mounted 450gal (2,045L) drop tank, photographed using a long exposure during a preliminary static-drop test. The pins extending from the pylon were for forcibly separating the tank from the pylon. BAE Systems via Brooklands Museum

  Details of the ventral tank fuel system. BAE Systems

  The ventral tank. Primarily intended for the ferry role, this large 1,425gal (6,480L) tank was unlikely to be used on combat missions. BAE Systems

  Hydraulic system

  Hydraulics were divided into two systems: general services (airbrakes, flaps, weapons bay doors, intake cones, intake auxiliary doors, artificial feel, wheel brakes and steering, flight refuelling probe, etc.), and flying controls (tailerons and flaps, fin, braking parachute release). Each system had a backup, so in total there were four self-contained systems on the aircraft. Each engine drove one controls and one services pump via the accessory drive gearbox, with pressure normally held at 4,000ps
i (280kg/ sq cm). Relief valves would vent excess if it rose above 4,800psi (337kg/sq cm).

  The layout of the hydraulic system. Two serious hydraulic leaks were experienced during flight testing, one near the end of Flight 23 as an artificial feel unit leaked, emptying the Services 1 system, and the fin-jack-ram fracture that caused Flight 25 to be aborted before take-off. BAE Systems via Brooklands Museum

  Electrical system components. BAE Systems via Brooklands Museum

  The two controls systems included accumulators that acted as boosters during high-rate manoeuvring and also provided limited control surface movement in the event of hydraulic failure (if both engines failed, for example). Windmilling engines would provide enough hydraulic pressure to maintain adequate control power as long as airspeed was kept above 220kt (250mph; 400km/h). Services system No. 1 included two accumulators, one for the artificial-feel units and yaw/roll gearbox and the other for the aft wheel brakes. Services No. 2 had three accumulators, the first again for artificial feel and yaw/roll gearbox, the second for the forward wheel brakes, and the last for AAPP raising/ lowering and emergency nosewheel steering.

  The layout of the electrical system. BAE Systems via Brooklands Museum

  A Rotax advert of March 1965. via Brooklands Museum

  The choice of DP47 (Silcodyne H) for the hydraulic fluid was made because of its superior properties at high temperatures and pressures, but gave rise to a succession of problems throughout the project, with filters, seals and couplings all proving inadequate at one point or another. Unlike normal hydraulic fluid, DP47 was colourless, and this led to an unexpected problem in that minor spills were difficult to spot, and the airframes gradually became increasingly contaminated with DP47 throughout production. An expensive clean-up was necessary before any paint could be applied to the first three aircraft, and measures had to be introduced to try to reduce contamination of subsequent airframes.

  Electrical system

  The aircraft’s electrical supply was designed to have safety factors similar to other systems, with duplication and backups. It was a 200V three-phase AC system, driven by two engine-gearbox-mounted 60KVA generators. Ground-running power for the engine starting system and certain other essential services was provided by an auxiliary generator driven by the AAPP, located in the AAPP bay just ahead of the weapons bay. The AAPP could not be run in flight, and before use it would be extended out of its bay. The door hinged at the forward end of the bay and the AAPP exhausted to the starboard side.

  There was a further hydraulically driven emergency generator in the Doppler bay, which, in the unlikely event of both generators failing, would provide power to the essential bus-bar and maintain the aircraft’s essential services. Transformer-rectifier units in the port and starboard equipment bays provided DC power. A battery in the starboard equipment bay also gave limited DC power for some systems (this was very limited on the first aircraft, the capacity and longevity of the battery coming in for sustained criticism from the crews).

  External DC and AC supplies could be plugged in to access points within the nose gear bay. For refuelling operations, only DC would be required.

  CHAPTER SEVEN

  The Engine

  Medway versus Olympus

  Both English Electric and Vickers had chosen the same Rolls-Royce engine for their GOR.339 submissions, the RB.142 Medway turbofan. This was to be a military derivative of the RB.141, an engine on which Rolls-Royce was working (and financing itself) for a medium-haul airliner for BEA. (This was the D.H.121, later to be known as the Trident, which ended up using an entirely different engine, the Rolls-Royce Spey.) Turbofans, or bypass engines, direct some of their compressed intake air around the central core of the engine, bypassing the combustion chambers. At the time, dealing with the differing flows of the slower and cooler bypass air and the faster and hotter exhaust air in the reheat jet pipe was thought by some to be a likely cause of development problems, and this, plus lower expected thrust from this type of engine, were factors that did not put it in a favourable position. One big plus point on the Medway’s side was its fuel consumption, which was going to be considerably better than that of turbojet designs.

  Among the alternatives was a well-known turbojet, the Bristol Aero Engines Olympus. By mid-1958 the basic Olympus design was already seven years old. It had been in service, powering RAF Vulcans, since 1956, and it was building a reputation for reliability and ruggedness unmatched by any similar engine. The Olympus Mk 101 (B.0l.1) was a 11,000lb (5,000kg)-thrust engine that powered the Vulcan B.1; the Mk 201 (B.0l.6) that was being delivered to power the Vulcan B.2 had been uprated to 17,000lb (8,000kg). An even more powerful engine, the Mk 301 (to be fitted to the Vulcan B.2 during the 1960s), was also under development. Bristol had used the 201 as the basis for a research engine, the B.0l.14R (R for reheat), with turbine and combustion chambers redesigned to operate at lower temperatures and thus improve specific fuel consumption (SFC). By May 1958 the 14R had undergone enough testing to demonstrate 13,620lb (6,180kg) dry thrust and considerable improvements in fuel economy compared with previous Olympus variants, and the B.0l.15R (a development of the 14R) was beginning testing. Both the 14R and 15R were proposed by Bristol as possible powerplants for the GOR.339, with reheat units based on work by an American company called Solar. Brochure figures for the 15R predicted a thrust of 16,400lb (7,443kg) (dry, 0.779 SFC) and 24,700lb (11,210kg) (reheat, 1.785 SFC). The Medway’s predicted figures were nearly identical in terms of thrust, with better fuel consumption in dry power and worse fuel consumption in reheat, but overall it appeared to be a better choice of engine.

  In July 1958 Bristol made a further submission based on a developed 15R, the B.Ol.22R, and withdrew the 14R from the running. This brochure predicted that the 22R would produce a thrust of 16,803lb (7,627kg) (dry, 0.71 SFC) and 30,554lb (13,868kg) (reheat, 1.77 SFC). Purely on brochure figures the Medway now lagged behind, if only slightly, in all but dry power fuel consumption. In September 1958 officials from the MoS and the Treasury made a decision that put Rolls-Royce out of the running, with only cursory attention paid to the actual performance of the competing engines. Rolls-Royce’s estimates on development cost for the engine, after appropriately pessimistic adjustments by the Ministry, were seen as much the same as those from Bristol Aero Engines, but unfortunately Rolls-Royce was simply doing too well as an engine company. Both the MoS and the Treasury felt that it was ‘desirable to keep some rival to Rolls-Royce’ in business. Rolls-Royce products were apparently becoming the engines of choice in both the civil and military fields, and they did not want them to become a monopoly. Bristol Aero Engines went away and came back with revised cost figures that drastically undercut Rolls-Royce. Furthermore, the men from the Ministry also believed that work on the 22R would directly benefit from the Olympus 21 (Mk 301) development and lead to a faster and cheaper result from both engines, despite there being no real commonality in the respective tasks each engine was destined to perform.

  A general-arrangement drawing of the Bristol-Siddeley Olympus B.0l.22R. Damien Burke

  Rolls-Royce’s proposals were discarded, and the MoS issued an edict in December 1958 that no further proposals for a powerplant would be sought from Rolls-Royce. The arrival of a final brochure from the company merely gave the MoS the opportunity to cite insufficient take-off thrust for the Medway as a reason to go with the Olympus. When the decision was made to go ahead with the GOR.339 (or TSR2 as it was to be known), the choice of engine was also to be announced. There was, technically, ‘not much in it’, the Ministry admitted privately, but the more important factors were the future size and shape of the aero-engine industry and the shotgun marriage that could be arranged between Bristol Aero Engines and Armstrong Siddeley Motors. This had been pushed into being even before the official announcement of the Olympus decision was made, and the result was Bristol Siddeley Engines, later Bristol Siddeley Engines Ltd.

  Unaware of the real reasons for the decision, BAC was not impressed by th
e choice, and George Edwards stated in a meeting with the Ministry on 1 January 1959 that they needed a further two months to assess the merits of the two engines, and that in any case such a decision should be a choice left to the aircraft designer. The Ministry was not about to let its decision be overturned, and wanted agreement that very day from BAC. After retiring to discuss it, the meeting reconvened and BAC submitted to the Ministry’s choice, on the condition that when the announcement was made it would be made clear that the responsibility for the selection of the engine lay with the Ministry and not with BAC.

  Rolls-Royce, unaware of the other factors behind the decision, produced a flurry of communications protesting against the choice of the Olympus, further detailing its belief in the superiority of the Medway, and attempting to enlist help from the RAF and the National Gas Turbine Establishment (NGTE) to press its case. It seemed pretty clear that the Medway was to be a more advanced design (and not ‘handicapped’ by being a development of an existing engine), would be more economical and had a promise of great development potential. Unfortunately for Rolls-Royce (and arguably, as it transpired, for the project as a whole) all of its protests came to naught.

  In February Rolls-Royce received a final brush-off from the MoS that made it clear, based on RAE and NGTE findings, that the Olympus was regarded as being more powerful and more suitable for the TSR2. Contract negotiations and technical discussions with BSEL rumbled on through the first half of 1959. In March the NGTE completed its evaluation of the 15R and found that Bristol’s fuel consumption figures had been optimistic to the tune of about 5 per cent, a massive difference which would have serious consequences for the amount of fuel needed and thus the weight of the aircraft. If a reduction in range could be accepted, however, the engine would do the job, especially if take-off thrust could be improved. The NGTE’s initial evaluations of the 22R figures indicated that fuel consumption on that engine was also likely to be worse than predicted, but, without an actual engine to run, further evaluation was going to be theoretical.